3 ms·

Ideally, never. Tying an open platform into features that exist only in one SaaS offering from an advertising company isn't a good way to keep it open.

azinman2 7y ago
Supporting the email provider of hundreds of millions of people with a feature that should be common for all is somehow bad for the future of Thunderbird? Not giving people what they want will drive them away, which is the worst way to keep something open. It’s not like you need proprietary binary blobs to support Gmail...
- dmos62 7y ago
  If you put gmail features into TB, it will become a gmail client, not an email client. The reason why all email servers and clients can talk to each other at the moment is because it's been well standardized, as opposed to, say, instant messaging. I'm all for improving email, but if done irresponsibly it will introduce division due to incompatibility and damage the ecosystem.
  - jacobsenscott 7y ago
    This is also the reason we can't have nice things using open standards, like labels. Once everything is tied to an ossified but open standard that's the end of the road as far as innovation is concerned.
    - dmos62 7y ago
      It's not about openness, but about federation.
    - JetSpiegel 7y ago
      Google could have tried to standardize IMAP2, but they didn't. Now only JMAP exists as a more modern mail standard.
selfishgene 7y ago
I understand your point, but gmail labels are such a good idea that Thunderbird should implement them as a native feature (even perhaps proposing it as extension of IMAP standard). Also think Thunderbird should implement JMAP (Fastmail standard) and/or Google RESTful gmail API because these are open standards worth supporting. Another feature that I would like to see more fully supported in Thunderbird (through better synchronization with mail provider server back-end) is editing emails, to include most importantly the ability to remove bulky attachments without losing email text or touching important meta-data. Unfortunately gmail treats each email as an immutable object, and the hacks that have been suggested to date to work around this limitation have all been clunky at best.
- bonzini 7y ago
  There's no need for an IMAP extension; there's already tags, but Thunderbird is pretty much the only mail agent that has some kind of support for IMAP tags. Google uses the folder hack in part because support for tags is so bad. However it would be nice if Google supported both folders and tags.
- omnimus 7y ago
  It seems like it's other way around. There is IMAP spec for keywords https://tools.ietf.org/html/rfc5788 https://tools.ietf.org/html/rfc5788 gmail just never implemented it (for whatever reason the spec is probably bad). I am afraid when we start adding gmail specific features at certain point google makes compatibility breaking change and becomes email. The people not using gmail will have to start because of network effect and email died.

azinman2 7y ago

Supporting the email provider of hundreds of millions of people with a feature that should be common for all is somehow bad for the future of Thunderbird? Not giving people what they want will drive them away, which is the worst way to keep something open. It’s not like you need proprietary binary blobs to support Gmail...

dmos62 7y ago
If you put gmail features into TB, it will become a gmail client, not an email client. The reason why all email servers and clients can talk to each other at the moment is because it's been well standardized, as opposed to, say, instant messaging. I'm all for improving email, but if done irresponsibly it will introduce division due to incompatibility and damage the ecosystem.
- jacobsenscott 7y ago
  This is also the reason we can't have nice things using open standards, like labels. Once everything is tied to an ossified but open standard that's the end of the road as far as innovation is concerned.
  - dmos62 7y ago
    It's not about openness, but about federation.
  - JetSpiegel 7y ago
    Google could have tried to standardize IMAP2, but they didn't. Now only JMAP exists as a more modern mail standard.

selfishgene 7y ago

I understand your point, but gmail labels are such a good idea that Thunderbird should implement them as a native feature (even perhaps proposing it as extension of IMAP standard). Also think Thunderbird should implement JMAP (Fastmail standard) and/or Google RESTful gmail API because these are open standards worth supporting. Another feature that I would like to see more fully supported in Thunderbird (through better synchronization with mail provider server back-end) is editing emails, to include most importantly the ability to remove bulky attachments without losing email text or touching important meta-data. Unfortunately gmail treats each email as an immutable object, and the hacks that have been suggested to date to work around this limitation have all been clunky at best.

bonzini 7y ago
There's no need for an IMAP extension; there's already tags, but Thunderbird is pretty much the only mail agent that has some kind of support for IMAP tags. Google uses the folder hack in part because support for tags is so bad. However it would be nice if Google supported both folders and tags.
omnimus 7y ago
It seems like it's other way around. There is IMAP spec for keywords https://tools.ietf.org/html/rfc5788 https://tools.ietf.org/html/rfc5788 gmail just never implemented it (for whatever reason the spec is probably bad). I am afraid when we start adding gmail specific features at certain point google makes compatibility breaking change and becomes email. The people not using gmail will have to start because of network effect and email died.